ContextLife

Inspiration

Every day, we have countless thoughts, conversations, and experiences — but most of them vanish. When we sit down with AI assistants like Claude, we have to re-explain who we are, what we're working on, and what happened today.

What it does

ContextLife was born from a simple idea: your life should be the context.

How we built it

• Swift native iOS app
• AVAudioEngine for background audio recording
• WhisperKit (Core ML) for on-device transcription
• Core Location (Significant-Change) for battery-efficient location tracking
• iCloud Drive for Obsidian sync
• RevenueCat for subscription management ## Challenges we ran into
• Apple's Speech Framework has strict background limitations — we pivoted to WhisperKit
• Balancing always-on recording with battery life
• Designing a "set and forget" UX that doesn't interrupt daily life ## Accomplishments that we're proud of
• Achieved zero API cost for transcription with on-device processing
• Full privacy — no audio ever leaves the device
• Seamless Obsidian integration for PKM users ## What we learned
• iOS background processing is tricky but manageable with the right architecture
• Separating recording from transcription is key for battery optimization
• The PKM community (Obsidian users) is hungry for AI-powered tools ## What's next for ContextLife
• Photo capture (manual + automatic)
• Claude Vision integration for visual context
• Smart summaries generated on-device
• Apple Watch companion app

Built With

• avaudioengine
• core-location
• core-ml
• icloud-drive
• ios
• revenuecat
• swift
• whisperkit